From: SINCLAIR  DOUGLAS N <sinclai@ecf.toronto.edu>
Date: Fri, 9 Dec 94 06:53:36 PST
To: cypherpunks@toad.com
Subject: Secure DAC?
Message-ID: <94Dec9.095259edt.3818@cannon.ecf.toronto.edu>
MIME-Version: 1.0
Content-Type: text/plain


-----BEGIN PGP SIGNED MESSAGE-----

I was looking through the PMI analog IC databook from 1988.  Specifically,
I was looking at the "DAC-08 8-bit high-speed multiplying D/A converter".
The chip is interesting in that it has two outputs.  One is the current
corresponding to the digital value on the input lines, and the other is
the current corresponding to NOT the digital values.  Thus, the two currents
sum to a constant.  The book says:

    "Power consuption may be calculated as follows:

    Pd = (I+) (V+) + (I-) (V-).  A useful feature of the DAC-08 design
    is that supply current is constand and independent of input logic states;
    this is useful in cryptographic applications and further serves to
                      ^^^^^^^^^^^^^
    reduce the size of the power supply bypass capacitors."
        -- Page 11-34.  All typos mine.

I fail to see how this would increase cryptographic potential.  However, 
it would seem to mask the device's EMR.  The application notes state that
it is useful in CRT display drivers, audio encoders, &c.  These are all
devices that are succeptable to TEMPEST monitoring (or whatever the
correct name is).  If one output goes to the desired next stage (the
electron gun in a CRT for example) and the other goes to a matched dummy
load, it would seem that the overall EMR is constant.  This approach
would obliviate the need for bulky shielding, though it might not be 
as secure.


-----BEGIN PGP SIGNATURE-----
Version: 2.6

iQCVAgUBLuhtmcyi36CIyUXVAQHPOgQAtfWIF40YQU+pYhYok734qzePhkDmmaaR
8WLXVB9UvZBi7psIuXRsXoKkODaQIJecQZ6UsIrEfBr1Lor+ZBe7e7fOpWiO5jkE
gMC+/a62z4xOyr4ukrEsZPG4WfHLR1SSob+CFla/JjiL2Hp3+I2wgtieY7KHTn0m
13Zpr/eayLM=
=F8pd
-----END PGP SIGNATURE-----